Elon Musk, the Australian government, and X: what's really going on

In April, the Australian eSafety commissioner, Julie Inman Grant, required both Meta and X to remove a video of a church stabbing. This was done under the Online Safety Act, part of Australian law. The material was deemed to be “class 1” material, which depicts gratuitous or offensive violence with a high degree of impact or detail.

It’s not the first time that this has been done - in the 2022-2023 financial year, the Australian government had issued three notices to overseas services about this type of material.

Both Meta and X agreed, and removed the posts.

Specifically, X had geo-blocked the 65 posts on its platform. In X’s eyes, X had complied with Australian law. People with an Australian IP address couldn’t see the posts.

However, the Australian government wasn’t entirely happy - because people with a VPN were able to see the posts, if they just connected to a VPN to send their traffic to another country. With a VPN, Australians were able to see this content. So, the eSafety Commissioner has asked for X to comply with the law.

This has been rather lazily reported as “the Australian government wants this content removed globally”. As far as I can see, it’s not asked that.

We all agree with Musk: except he’s dishonest

Elon Musk hasn’t really helped this discussion (when does he ever?) by recommending that you download VPNs to evade country bans.

In a tweet on April 23, he wrote:

“Our concern is that if ANY country is allowed to censor content for ALL countries, which is what the Australian “eSafety Commissar” is demanding, then what is to stop any country from controlling the entire Internet?”

That’s absolutely right, though, she’s the Commissioner, not the Commissar. But we all agree with that thought. It’s a lovely thought, but it’s also entirely irrelevant.

It’s absolutely not what Australia is asking for. It’s a dishonest, deceptive tweet from Musk, because the Australian court hasn’t asked for this content to be removed globally. It has no right to do that, and that’s massive overreach.

Here’s all Australia is asking for

If you agree that this content should be removed from Australian view - that’s a big “if”, but we’ll assume that our benevolent leaders are correct in requiring its removal for now, here’s what should be happening.

The Australian court is saying: X should make best efforts to ensure that the content is not visible to people in Australia, so as to follow the requirements of the Online Safety Act.

X hasn’t done that. It’s merely removed the content from view when connecting from an Australian IP address, rather than removing the content from view from Australians.

They are most certainly not doing “best efforts”.

“Best efforts”

Back in 2021, I requested my data from Twitter, to discover what the company knew about me. They know an awful lot.

Particularly, the ip-audit file, when I requested my data, showed 2,072 IP addresses that I’d connected to Twitter from. That’s 60 days of connection data.

That 60 days of IP connection data should be relatively clear that I’m based in Australia. Even if I do connect via a VPN to Twitter, it should realise that over 80% of my connections have been fron Australia, and that there is a high degree of certainty that I’m an Australian, and so therefore it should deny the banned content.

On Android, at least, X has a number of signals to suggest that I’m based in Australia. It can see my bank app. For many people, it has GPS location data (I’ve turned that off). It can see the ads it’s displayed to me, and my interactions with them. It knows who I follow, which is also a fine set of signals to suggest that I may be an Australian.

Instead, X hasn’t bothered using any of these signals to work out if I’m Australian - using a blunt instrument of “my current IP address”.

Where Musk is based is irrelevant

“But X isn’t based in Australia, so the Australian government has no power over it” isn’t the right argument here. The Australian government has powers to protect Australians (in this case) from content available in Australia.

The Australian government doesn’t, though, have power over what Norwegians, Canadians, or the Japanese see on the internet.

Language is important here. It should be clear in its demands to X that it requires best efforts that this content is not visible to people in Australia. As the law says.

As it is, the Australian government is accused, by lazy media reporting, of wanting to censor the internet globally. That’s really not the case. They’re just wanting X to use technology already available to it to properly block this content.

We’ll discover more, when another court hearing happens in a week; though my suspicion is that it’ll be seen as an embarrassing climbdown by the Australian government. It won’t be - it’ll just be a request to X to do their job.